Working Paper: NBER ID: w20298
Authors: Lorenz Kueng; Evgeny Yakovlev
Abstract: We study the long-run effects of Russia's anti-alcohol campaign, which dramatically altered the relative supply of hard and light alcohol in the late 1980s. We find that this policy shifted young men's long-run preferences from hard to light alcohol decades later and we estimate the age at which consumers form their tastes. We show that the large beer market expansion in the late 1990s had similar effects on young consumers' tastes, while older consumers' tastes remained largely unchanged. We then link these long-run changes in alcohol consumption patterns to changes in male mortality. The shift from hard to light alcohol reduced incidences of binge drinking substantially, leading to fewer alcohol- related deaths. We conclude that the resulting large cohort differences in current alcohol consumption shares explain a significant part of the recent decrease in male mortality. Simulations suggest that mortality will continue to decrease by another 23% over the next twenty years due to persistent changes in consumer tastes. Program impact evaluations that focus only on contemporaneous effects can therefore severely underestimate the total effect of such public policies that change preferences for goods.
Keywords: public policy; alcohol consumption; mortality; Russia; anti-alcohol campaign
JEL Codes: D12; E21; G02; I10
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Long-term changes in consumer preferences (D12) | Predicting a decline in future mortality rates (J11) |
| Anti-alcohol campaign (L66) | Reducing mortality due to alcohol poisoning and other causes (I12) |
| Anti-alcohol campaign in the late 1980s (L66) | Shifted preferences from hard to light alcohol (L66) |
| Shifted preferences from hard to light alcohol (L66) | Reduced binge drinking (I12) |
| Reduced binge drinking (I12) | Decreased alcohol-related deaths (I12) |
| Higher vodka consumption (L66) | Increased male mortality rates (J11) |
| Type of alcohol consumed (L66) | Impacts mortality independent of total amount consumed (I12) |